INDEX

Explanations

arguments and counterarguments in legal or political contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

opposing arguments or claims

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 autorytatywna

-0.74

parsedMessage

-0.68

 preſent

-0.58

Personensuche

-0.54

 houſe

-0.53

principalColumn

-0.52



-0.51

thâu

-0.50

 Numerade

-0.50

 Houſe

-0.49

POSITIVE LOGITS

disposing

0.34

Bioaccumulative

0.33

 Allociné

0.33

LookAndFeel

0.32

 others

0.32

 wensen

0.31

նակ

0.30

则

0.30

qui

0.29

ționale

0.29

Activations Density 0.078%

arguments and counterarguments in legal or political contexts

opposing arguments or claims

No Comments

No Known Activations